Citation: Cell Death and Disease (2016) 7, e2082; doi:10.1038/cddis.2016.8 OPEN & 2016 Macmillan Publishers Limited All rights reserved 2041-4889/16 www.nature.com/cddis ANP32B deficiency impairs proliferation and suppresses tumor progression by regulating AKT phosphorylation

S Yang1,6, L Zhou2,6, PT Reilly3, S-M Shen1,PHe1, X-N Zhu1, C-X Li1, L-S Wang4, TW Mak5, G-Q Chen*,1 and Y Yu*,1

The acidic leucine-rich nuclear phosphoprotein 32B (ANP32B) is reported to impact normal development, with Anp32b-knockout mice exhibiting smaller size and premature aging. However, its cellular and molecular mechanisms, especially its potential roles in tumorigenesis, remain largely unclear. Here, we utilize 'knockout' models, RNAi silencing and clinical cohorts to more closely investigate the role of this enigmatic factor in cell proliferation and cancer phenotypes. We report that, compared with Anp32b wild- type (Anp32b+/+) littermates, a broad panel of tissues in Anp32b-deficient (Anp32b− / −) mice are demonstrated hypoplasia. Anp32b− / − mouse embryo fibroblast cell has a slower proliferation, even after oncogenic immortalization. ANP32B knockdown also significantly inhibits in vitro and in vivo growth of cancer cells by inducing G1 arrest. In line with this, ANP32B protein has higher expression in malignant tissues than adjacent normal tissues from a cohort of breast cancer patients, and its expression level positively correlates with their histopathological grades. Moreover, ANP32B deficiency downregulates AKT phosphorylation, which involves its regulating effect on cell growth. Collectively, our findings suggest that ANP32B is an oncogene and a potential therapeutic target for breast cancer treatment. Cell Death and Disease (2016) 7, e2082; doi:10.1038/cddis.2016.8; published online 4 February 2016

The acidic leucine-rich nuclear phosphoprotein 32 kDa lethality and reduced body weight,22–25 indicating a greater (ANP32) protein family are characterized by a N-terminal importance of Anp32b in normal development. In addition, leucine-rich repeat domain and a C-terminal low-complexity gene expression analysis indicates that elevated ANP32B acidic region.1 In mammals, the ANP32 family has at least mRNA expression correlates with highly proliferative three members named ANP32A, ANP32B and ANP32E, tissues.22 We also showed that ANP32B acts as a negative and they regulate a wide spectrum of biological processes regulator of leukemic cell apoptosis,8 and inhibits all-trans – – including chromatin regulation,2 6 caspase activation,7 9 retinoic acid induced leukemic cell differentiation.26,27 protein phosphatase inhibition10–12 and intracellular Although these studies strongly suggested ANP32B as a transport.13,14 Although early investigations suggested that master regulator of cell fate determination, its cellular and three ANP32 members functionally overlap,10 they are molecular mechanisms are still not understood. Considering reported to have diverse roles in cancer progression. ANP32A that some physiological and pathological processes share was shown to inhibit cell transformation15–17 and has reduced many common molecular regulators,28 and ANP32B mRNA expression in prostate and breast cancer.18,19 ANP32E was expression is a marker for aggressive breast cancer,22 we reported to have enhanced expression in gastric cancer,20 and proposed that ANP32B also functions in breast cancer. Here, a high expression of ANP32E was associated with better we used Anp32b-knockout mice, multiple breast cancer cell survival rate in follicular lymphoma.21 Previously we reported lines and clinical patient samples to uncover the potential role that ANP32B, also designated as PHAPI2 or SSP29, is a for ANP32B in cell proliferation of both mouse embryo negative prognostic indicator for human breast cancer.22 Full fibroblasts (MEFs) and breast cancer cells, and find that loss analysis of the expression and functional role of ANP32B in of ANP32B by knockout or RNAi silencing reduced rates of cell cancer progression has still not been undertaken. proliferation. We also show that RNAi silencing induces an Knockout mouse studies demonstrated that loss of Anp32b, extended G1-phase of the cell cycle. In addition, phosphoryla- but not Anp32a and Anp32e, caused a high degree of perinatal tion of AKT, an upstream regulator of cell cycle-associated

1Key Laboratory of Cell Differentiation and Apoptosis of Chinese Ministry of Education, Rui-Jin Hospital, Shanghai Jiao-Tong University School of Medicine (SJTU-SM), Shanghai, China; 2Department of Surgery, Branch of Shanghai First People’s Hospital, SJTU-SM, Shanghai, China; 3Laboratory of Inflammation Biology, National Cancer Centre Singapore, Singapore, Singapore; 4State Key Laboratory of Genetic Engineering, Minhang Hospital, Fudan University, Shanghai, China and 5Campbell Family Cancer Research Institute, University Health Network, Toronto, ON, Canada *Corresponding author: G-Q Chen or Y Yu, Department of Pathophysiology, Shanghai Jiaotong University School of Medicine, 280, Chong-qing South Road, Shanghai 200025, China. Tel: +86 21 63846590 776573; Fax: +86 21 64154900; E-mail: [email protected] or [email protected] 6These two authors equally contributed to this work. Abbreviations: ANP32, acidic leucine-rich nuclear phosphoprotein 32; BLI, bioluminescence imaging; CCK-8, cell counting kit-8; CDK, cyclin-dependent kinase; DMBA, 7,12-dimethylbenz(a)anthracene; MEFs, mouse embryo fibroblasts; FBS, fetal bovine serum; GFP, green fluorescence protein; IHC, immunohistochemical staining; IRS, immunoreactive score; PHAPI2, putative HLA-DR-associated protein I-2; PHLPP, PH domain leucine-rich repeat protein phosphatase; PDK1, pyruvate dehydrogenase kinase isozyme 1; PP2A, protein phosphatase 2A; PTEN, phosphatase and tensin homolog; SSP29, silver-stainable protein 29 Received 02.10.15; revised 31.12.16; accepted 04.1.16; Edited by J Chipuk ANP32B deficiency suppresses proliferation and tumorigenesis S Yang et al 2

proteins, is lower coincident with reduced ANP32B upon homozygous Anp32b− / − mice,22 the Balb/c-congenic silencing and in both mouse and human cancers. Anp32b− / − mice also had a statistically significant reduction of body weight compared with wild-type (Anp32b+/+) and +/ − Results heterozygous (Anp32b ) mice at 4 weeks after birth (Figure 1a). Here we also found the decreased weigh of the − − Anp32b− / − MEFs are impaired in cell proliferation and Anp32b / mice was accompanied by the reduced size of oncogenic transformation. As seen in mixed-bred organs such as the heart, liver, spleen and kidney (Figures 1b

Figure 1 Anp32b deficiency impairs normal cell proliferation and oncogenic transformation. (a) The body weight of 22 Anp32b +/+,30Anp32b +/ − and 8 Anp32b − / − mice at 4 weeks of age. Data were analyzed using Mann–Whitney U-test. **Po0.01. (b) Photographs of appearance and organs of Anp32b− / − mice and Anp32b+/+ littermate at 4 months of age. (c) The weight of organs from Anp32b+/+ and Anp32b− / − mice at 4 months of age. Data are presented as mean ± S.D. and significance is *Po0.05 (n = 4). (d) Western blots for the indicated protein in primary Anp32b+/+ and Anp32b− / − MEF cells. (e) Proliferation of primary Anp32b+/+ and Anp32b− / − MEFs was monitored for the indicated times. Nt/N0 represents the cell number at a given time normalized to the cell number at day 0. Data are presented as mean ± S.D. The experiment is representative of three separate experiments. (f) Western blots for the indicated protein in immortalized Anp32b+/+ and Anp32b− / − MEF cells. (g) Proliferation of immortalized Anp32b+/+ and − − Anp32b / MEFs was monitored for the indicated times. Data are presented as mean ± S.D. (h) Immortalized MEF cells were subcutaneously injected into nude mice and the size of masses were analyzed. Data are presented as mean ± S.D.

Cell Death and Disease ANP32B deficiency suppresses proliferation and tumorigenesis S Yang et al 3 and c). To determine whether this is due to reduced cell ANP32B promotes cell cycle progression. To examine volumes or hypoplasia, histological H&E staining and flow whether the effect of ANP32B on breast cancer cell cytometric evaluation were performed. The results showed proliferation is partly due to cell cycle arrest, we used a that there was no obvious cell size difference between these double thymidine block to synchronize cells at the G1/S organs in Anp32b− / − and Anp32b+/+ mice (Supplementary border, followed by addition of nocodazole to block cells in Figure S1A and B). However, cell numbers in organs such G2/M. Flow cytometry analysis was then used to monitor the as spleen and thymus were dramatically decreased in progression of cells from G1/S to G2/M. The results showed Anp32b− / − mice (Supplementary Figure S1C), supporting that after nocodazole treatment within 3 h 89.5% of cells in that Anp32b deficiency causes a hypoplastic phenotype in shNC BT549 cells entered the S phase, whereas cells at the multiple organs. S phase only had 59.9 and 63.8%, respectively, in sh32b#1 To functionally characterize the role of Anp32b in normal and sh32b#2-infected cells. After nocodazole treatment for +/+ cell proliferation, we isolated MEFs from Anp32b and 9 h, only 46.3 and 39.8% of total cells entered the G2/M − − Anp32b / mice. The ANP32B protein was totally knocked phase in two sh32b-infected BT549 cells compared with out in Anp32b− / − MEF cells (Figure 1d). Cell proliferation 69.6% of total cells in shNC BT549 clones (Figure 3a). In assay showed that primary Anp32b− / − MEF cells divided addition, we analyzed the cell cycle regulatory proteins. As significantly more slowly than Anp32b+/+ MEF cells shown in Figure 3b, cyclin D1 protein level had no alteration (Figure 1e). Given the effect of Anp32b on cell proliferation, between NC and sh32b-infected cells. However, cyclin D1 we set out to assess whether Anp32b deficiency could inhibit was time-dependently increased in shNC-infected BT549 oncogenic transformation. To this end, MEFs were immorta- cells upon nocodazole treatment, which was remarkably lized by infection with a retrovirus encoding two oncogenes, inhibited in sh32b-infected cells. Similarly, we re-introduced adenovirus 5 E1A and constitutively active form of H-RasV12 ANP32B into sh32B#2-transfected BT549 cells (Figure 3c). (Figure 1f). The results showed that the immortalized Cell cycle analysis showed that complementation by ANP32B − / − Anp32b MEF cells also presented lower cell proliferation could rescue the cell cycle G1 phase arrest in sh32b BT549 than Anp32b+/+ MEF cells (Figure 1g). We also subcuta- cells (Figure 3d). Collectively, these results indicate that +/+ − / − neously injected immortalized Anp32b and Anp32b ANP32B promotes cell cycle progression at the G1 phase. MEF cells into nude mice. Mice xenografted with Anp32b− / − MEF cells developed significant smaller tumors compared Loss of ANP32B suppresses breast tumor growth with the mice injected with Anp32b+/+ MEF cells (Figure 1h). in vivo. These in vitro results prompted us to examine All these data suggest a functional role of Anp32b in the whether ANP32B has some effects on breast tumor growth proliferation of normal and transformed cells. in vivo. Hence, shNC, sh32b#2 and sh32b#2/GFP-ANP32B- infected MDA-231-D3H2LN cell line (Figure 2e), which was ANP32B knockdown inhibits breast cancer cell prolifera- derived from breast cancer cell line MDA-MB-231 with stable tion in vitro. We further investigated whether ANP32B luciferase expression, were injected into the mammary fat regulates cancer cell proliferation with breast cancer cells pad of nude mice. Luciferase photon fluxes were monitored as models. For this purpose, we used two pairs of shRNAs and the representative tumors are shown (Figures 4a and b). (sh32b#1 and sh32b#2) specifically against ANP32B to Consistent with the in vitro findings, reduction of ANP32B led generate stable ANP32B knockdown along with a control to a significant inhibition of tumor size at the fourth week after shRNA transfectant (shNC) in BT549, MCF7 and MDA-231- injection, which could be partially reversed by re-expression D3H2LN breast cancer cell lines. These two specific shRNAs of ANP32B (Figures 4a and b). Furthermore, ANP32B could effectively knockdown ANP32B but not its closely knockdown tumor cells showed obviously weaker Ki-67 related ANP32A expression in these breast cancer cell lines staining compared with control tumor cells, which could be (Figure 2a and Supplementary Figure S2A). Then, we also reversed by re-expression of ANP32B, suggesting that examined the effect of ANP32B knockdown on breast cancer ANP32B knockdown indeed decreased cell proliferation cell proliferation. As shown in Figures 2b and c, ANP32B in vivo (Figure 4c). All these data strongly suggested that knockdown significantly inhibited the growth of BT549 cells specific loss of ANP32B could significantly inhibit breast with no effect on their viability. Similar effects could also be cancer growth in vivo. seen in MDA-231-D3H2LN (Figure 2b) and MCF7 cells (Supplementary Figure S2B and C). Compared with the ANP32B is highly expressed in human breast cancer. control cells, in addition, BT549 and MCF7 cells with ANP32B Previously, we examined the relationship between ANP32B silencing showed markedly decreased colony formation mRNA expression and breast cancer patient prognosis using ability with reduced colony number and size (Po0.05; information from three available data sets and reported that Figure 2d and Supplementary Figure S2D and E). To patients whose tumors showed the highest ANP32B mRNA demonstrate that cell growth inhibition is specifically due to levels had significant shorter survival.22 Here we performed the silencing of ANP32B, we re-introduced GFP-tagged Immunohistochemical staining (IHC) staining on 50 breast ANP32B into sh32b#2-transfected MDA-231-D3H2LN cells, tumor tissues and the matched adjacent normal tissues, and found that re-expression of ANP32B could reverse and found that breast tumor tissues presented higher ANP32B knockdown-induced cell growth inhibition ANP32B expression compared with adjacent normal tissues (Figures 2e and g). Taken together, these data suggest that (Figures 5a and b). Moreover, an increase of ANP32B protein ANP32B may be closely associated with the proliferation of level in tumor tissues over adjacent normal tissues was also breast cancer cell lines. confirmed by western blot analysis in five paired clinical

Cell Death and Disease ANP32B deficiency suppresses proliferation and tumorigenesis S Yang et al 4

Figure 2 Knockdown of ANP32B inhibits breast cancer cells proliferation. (a) Breast cancer BT549, MDA-231-D3H2LN cells were stably infected with shNC and sh32b, and the indicated proteins were detected by western blot with β-actin as a loading control. (b) Cell counting of shNC- and sh32b-infected BT549, MDA-231-D3H2LN cells after 2, 4 and 6 days of growth. (c) Cell viability after 6 days of growth was measured by trypan blue exclusion. Data are presented as mean ± S.D. of triplicate in an independent experiment, which was repeated for more than three times. (d) The morphology of shNC- and sh32b-infected BT549 cells under phase contrast microscopy (upper). Influence of ANP32B on colony formation of BT549 cells. Representative dishes are presented (middle). The number and size of clones were calculated for each well of six-well plates and shown in the y axis in the bottom panel. Data are presented as mean ± S.D. and significance is *Po0.05, **Po0.01, which was repeated for more than three times. (e) ShNC- and sh32b- infected breast cancer MDA-231-D3H2LN cells were stably transfected with empty vector (EV) and GFP-tagged ANP32B, followed by immunoblots for the indicated proteins. (f) Cell counting of shNC/EV, sh32b/EV and sh32b/GFP-ANP32B MDA-231-D3H2LN cells after 3 days of growth. Data are presented as mean ± S.D. and significance is **Po0.01, which was repeated for more than three times. (g) Representative images from the morphology and colony formation of shNC/EV,sh32b/EVand sh32b/GFP-ANP32B MDA-231-D3H2LN cells

breast cancer specimens (Figure 5c). These data indicate ANP32B protein expression in breast cancer is directly related that ANP32B expression is enhanced in human breast with histological grade of cancer tissues. cancer at the protein level. We next evaluated the correlation between ANP32B ANP32B has positive correlation with p-AKT and regu- expression and clinicopathological parameters. As presented lates AKT activation. We analyzed the expressions of in Supplementary Figure S3, there was no significant cyclins such as cyclin D1/3, cyclin-dependent kinases correction for ANP32B expression with age or clinical stage (CDKs) including CDK4, CDK6, CDK2, CDK inhibitor p27, of breast cancer patients. However, ANP32B was associated as well as ERK and P38 in ANP32B silencing BT549 and significantly with histological grade. Higher levels of ANP32B MDA-231-D3H2LN cells. The results showed that knockdown was correlated with higher histological grade (I versus II; of ANP32B failed to change all these protein levels P = 0.0182, II versus III; P = 0.0231) (Figure 5d). Figure 5e (Supplementary Figure S4). More interestingly, ANP32B depicts three representative IHC images respectively for low, knockdown significantly reduced the phosphorylated AKT at medium and high ANP32B expressions of cancer tissues with Ser473 rather than AKT protein (Figure 6a). Of note, it did not different histological grade. These data suggest that elevated change phosphorylated ERK and P38 (Supplementary

Cell Death and Disease ANP32B deficiency suppresses proliferation and tumorigenesis S Yang et al 5

Double Nocodazle shNC sh32b#2 conthymidine 3h 6h 9h EV EV GFP-32b GFP-ANP32B shNC

sh32b#1

ANP32B

sh32b#2 β-actin

100 G2/M S Double Nocodazle thymidine G1 con 3h 6h 9h 50

phases ( %) shNC/EV Cells at various

0

sh32b#2/EV shNC sh32b#1 sh32b#2

Nocodazle Nocodazle Nocodazle con 3h 6h 9h con 3h 6h 9h con 3h 6h 9h sh32b#2/GFP-32b

ANP32B 100 G2/M cyclin D1 S 50 G1

1.0 0.98 1.12 1.38 2.55 1.03 0.94 1.12 1.13 1.22 0.98 1.02 1.08 1.10 1.16 phases ( %) Cells at Various 0 β-actin

Figure 3 ANP32B deficiency induces cell cycle G1/S arrest. (a) ShNC- and sh32b-infected BT549 cells were pretreated with thymidine twice and then treated with nocodazole for indicated times. DNA content of treated cells was analyzed by flow cytometry. (b) Equal amounts of the corresponding cell lysates were blotted for ANP32B, cyclin D1 and β-actin. (c) ShNC- and sh32b-infected breast cancer BT549 cells were stably transfected with empty vector (EV) and GFP-tagged ANP32B, followed by immunoblots for the indicated proteins. (d) ShNC/EV, sh32b/EV and sh32b/GFP-ANP32B BT549 cells were pretreated with thymidine twice and then treated with nocodazole for indicated times. DNA content of treated cells was analyzed by flow cytometry

Figure S4). A similar impact of ANP32B on AKT phos- BT549 cells were stably co-transfected with shNC or sh32b#2 − − phorylation was evident in Anp32b / MEF cells (Figure 6b). together with vector or flag-AKT, and the results showed that In line with this, a carcinogen 7,12-dimethylbenz(a)anthra- the p-AKT level was increased but still lower in sh32b cells cene (DMBA)-induced mammary tumors30 derived from compared with shNC BT549 cells (Figure 6g). As expected, − − Anp32b / but not Anp32b +/+ mice also displayed negative the ectopically expressed AKT could rescue ANP32B p-AKT staining expression (Figure 6c). knockdown-induced cell growth inhibition in BT549 cells We detected ANP32B and p-AKT expression by IHC (Figure 6h). Considering that Akt overexpression-restored staining in tumor tissues from breast cancer patients. pAkt levels might be responsible for reversion of effects in Figure 6d depicts two representative IHC images for ANP32B ANP32B knockdown cells, the HA-myr-AKT with constitutive and p-AKT expression. To better understand the correlation of activation of AKT31 was re-expressed in shNC and sh32b#2 ANP32B and p-AKT, we stratified the cohort into two groups BT549 cells (Figure 6i). Consistent with data in Figure 6h, the based on ANP32B staining (ANP32Blow and ANP32Bhigh), enforced expression of HA-myr-AKT could also rescue and P-AKT was scored using either median expression score ANP32B knockdown-induced cell growth inhibition in BT549 (Figure 5e) or the percentage of low and high expression cells (Figure 6j). All these results indicated that AKT signaling scores (Figure 5f). The results demonstrated a highly positive mediated ANP32B knockdown-induced cell growth inhibition. correlation between the ANP32B and P-AKT. Discussion ANP32B regulates breast cancer cell proliferation through AKT activation. To further investigate the contri- In this study, we used three models to examine the role of bution of AKT signaling to the role of ANP32B in cell ANP32B in cell proliferation and oncogenesis. The knockout proliferation, we ectopically expressed AKT in ANP32B mouse model demonstrated that ANP32B has a broad impact knockdown cells to evaluate whether it might overcome the on cell proliferation evidenced by hypoplasia in many organs, suppression effect of ANP32B deficiency on cell proliferation. and that loss of ANP32B inhibits normal cell proliferation and

Cell Death and Disease ANP32B deficiency suppresses proliferation and tumorigenesis S Yang et al 6

shNC/EV

sh32b#2/EV

sh32b#2/GFP-32b

1.00E+09 ∗ 8.00E+08 6.00E+08 shNC/EV sh32b#2/EV sh32b#2/GFP-32b 4.00E+08 ∗ Total flux Total 2.00E+08 0.00E+00 Ki67

Figure 4 Re-expression of ANP32B promotes breast cancer cell growth in vivo. MDA-231-D3H2LN cells infected with shNC/EV, sh32b/EV and sh32b/GFP-ANP32B were injected into right side of the mammary fat pad of the mice. (a) After 4 weeks, the tumor growth was monitored using bioluminescent imaging and the representative bioluminescent images (top)/total flux (bottom) are shown. Data are presented as mean ± S.D. and symbol *indicates Po0.05 between lined groups. (b) The macroscopic appearances of tumors at 4 weeks after injection. (c) Representative images of immunohistochemical staining for Ki-67 of mammary tumors at 4 weeks after injection

Figure 5 ANP32B expression is elevated in breast cancer tumors and positive correlates with historical grade of breast cancers. (a) ANP32B expression was plotted using the immunohistochemical scores as described in the Material and Methods. ANP32B expression scores are shown as box plots, with the horizontal lines representing the median; the bottom and top of the boxes representing the 25th and 75th percentiles, respectively; and vertical bars representing the range of data. We compared breast cancer tumors with matched adjacent normal breast epithelium using the Mann–Whitney test, n = 100. (b) Representative images from immunohistochemical staining of ANP32B from one pair of breast cancer and adjacent normal tissues. The scale bar represents 30 μm. (c) Expression of ANP32B in five pairs of clinical breast cancer specimens. N and T mean adjacent normal tissue and paired breast cancer tumor, respectively. (d) Box plots of ANP32B expression in breast cancers with different historical grades. Data were analyzed by one-way ANOVA test. (e) Representative images from immunohistochemical staining of ANP32B from three cases in different histological grades (1–3). The scale bar represents 30 μm

Cell Death and Disease ANP32B deficiency suppresses proliferation and tumorigenesis S Yang et al 7

Figure 6 The effects of ANP32B on the AKT activation and the correlation of ANP32B and p-AKT expression in breast cancer patients. (a) The expression of AKT and the phosphorylation of AKT in shNC- and sh32b-infected BT549 cells. (b) The expression of AKTand the phosphorylation of AKT in Anp32b+/+ and Anp32b − / − MEF cells. (c) H&E staining and immunohistochemical analysis were used to determine the level of phosphorylation of AKTand Ki-67 expression in mammary tumors from DMBA-induced Anp32b+/+ and Anp32b− / − mice. (d) Representative IHC images of breast cancer samples for the indicated proteins. The scale bar represents 30 μm. (e–f) Box plots of p-AKT scores (e) and the percentage of tumors with high and low p-AKTexpressions (f) in those with high and low ANP32B expressions. (g) ShNC- and sh32b-infected breast cancer BT549 cells were stably transfected with empty vector (EV) and Flag-AKT, followed by immunoblots for the indicated proteins. (h) ShNC- and sh32b-infected breast cancer BT549 cells were stably transfected with empty vector (EV) and HA-myr-AKT, followed by immunoblots for the indicated proteins. (i) Cell counting of EV- and Flag-AKT-transfected BT549 cells after 3 days of growth. Data are presented as mean ± S.D. and significance is **Po0.01, which was repeated for more than three times. (j) Cell counting of EV- and HA-myr-AKT- transfected BT549 cells after 3 days of growth. Data are presented as mean ± S.D. and significance is **Po0.01, which was repeated for more than three times

suppresses transformation. ANP32B silencing by RNAi also Clinical data set analyses showed that ANP32B protein inhibited breast cancer cell proliferation in vitro and in vivo. level is highly expressed in breast cancer patients and the Thus, ANP32B is an important proliferation-related nuclear elevated ANP32B protein expression is directly related with protein. Our further investigation with synchronize cells at the histological grade of breast cancer tissues. These data G1/S border, followed by addition of nocodazole to block cells suggested that ANP32B acts as a predictive indicator in in G2/M showed that ANP32B silencing significantly retarded breast cancer treatment. However, owing to the limitation of the progression of cells from G1/S to G2/M. patients sample information, relationships between ANP32B

Cell Death and Disease ANP32B deficiency suppresses proliferation and tumorigenesis S Yang et al 8

overexpression and clinical prognosis were not fully analyzed. IHC. The protein expression levels of ANP32B and p-AKT were analyzed by IHC Increased ANP32B in tumors and knockdown models also with anti-ANP32B and anti-p-AKT polyclonal antibody. All of the staining was correlated with high p-AKT expression, indicating a possible assessed by pathologists who were blinded to the origin of the samples and subject outcome. Each specimen was assigned a score according to the intensity of the mechanism through which ANP32B exerts its effect on cell nucleus, cytoplasmic and/or membrane staining (no staining = 0; weak staining = 1, proliferation and tumor progression. moderate staining = 2, strong staining = 3) and the area extent of stained cells The activated AKT pathway has been demonstrated to have (0% = 0, 1–24% = 1, 25–49% = 2, 50–74% = 3, 75–100% = 4). The final an essential role in normal cell and breast cancer cell immunoreactive score (IRS) was determined by multiplying the intensity score – proliferation.32 36 We found that the p-AKT level was with the extent score of stained cells, ranging from 0 (the minimum score) to 12 (the significantly decreased in ANP32B knockdown cells. Further- maximum score). Scores of ANP32B and p-AKT were divided into two classifications: low (IRS ⩽ 6) and high (IRS46). Detailed information of the two more, the restoration of AKT or constitutively active AKT cohorts was shown in Supplementary Table S3. expression could rescue the inhibition of cell proliferation by ANP32B deficiency, suggesting the inhibition of cell prolifera- Plasmids, siRNA designs and transfections. Human ANP32B cDNA tion by ANP32B deficiency is primarily mediated through AKT was cloned and inserted into pBabepuro Vector (Clontech, Mountain View, CA, activation in breast cancer cells. How ANP32B might regulate USA) with GFP tag. Two pairs of complementary siRNA oligonucleotides against the AKT activation is still unknown. Previous studies have ANP32B and a pair of scrambled negative control siRNA were synthesized by Invitrogen (Carlsbad, CA, USA), annealed and ligated into pSIREN-RetroQ vector reported that AKT activation could be regulated by many (Clontech). The target sequences for ANP32B were 5′-TGACTACCGAGAGAG genes, including PH domain leucine-rich repeat protein TGTC-3′ and 5′-GCGAAATAAACAGTTACTC-3′. Constitutively active AKT (HA-myr- phosphatase,37 serine/threonine protein phosphatase 2A AKT) and Flag-AKT were a kind gift from Dr. Yu Jianxiu in Shanghai Jiao-Tong (PP2A),38 pyruvate dehydrogenase kinase, isozyme 1 University School of Medicine. Retrovirus was generated by transient transfection of (ref. 39) and phosphatase and tensin homolog (PTEN),40 the 293T cell line with FuGENE9 transfection reagent (Roche, Basel, Switzerland). but our preliminary experiments showed that ANP32B failed After 48 h of transfection, the viral supernatant was harvested and used for infection of target cells. Stable retroviral transduction was achieved by infection for 48 h, after to interact with PP2A and PTEN (data not shown). So future which selection with either puromycin (1.5 μg/ml) was initiated. Selection was experiments will be needed to investigate the detailed stopped as soon as the non-infected control cell died off, and the media were mechanism about how ANP32B regulates AKT activation. replaced with normal-growing media. Totally, our results concluded that ANP32B, through its positive regulation of p-AKT, serves as a master enforcer of Western blots. For the protein expression analysis, standard western blotting cell proliferation. In the physiological context, knockout of was carried out with the following antibodies used: Rabbit polyclonal antibodies against phospho-AKT(Ser473), AKT, phospho-Rb(S780), ERK1, phospho-ERK1, ANP32B impedes the proper mammalian development, phospho-P38, vinculin, rabbit monoclonal antibodies against Cyclin D1, CDK2, whereas in the pathological context, ANP32B deficiency CDK4, P27 and mouse monoclonal antibodies against, CyclinD3, CDK6, β-actin functions as a suppressor of tumor growth and transformation. (Cell Signaling, Beverly, MA, USA), goat antibody against ANP32A, P38 (Santa Notably, ANP32B has been highly detected in breast cancer Cruz Biotech, Santa Cruz, CA, USA), rabbit antibody against ANP32B (Proteintech patients, thus highlighting ANP32B as a potential therapeutic Group, Chicago, IL, USA), E1A polyclonal antibody (Abcam, Cambridgeshire, UK) and H-Ras polyclonal antibody (Signalway, College Park, MD, USA). target for breast cancer treatment. Cell proliferation and colony formation assay. MEF cells proliferation was evaluated by the CCK-8 assay (WST-8; Cell counting kit-8 from Dojindo, Materials and Methods Kumamoto, Japan). In brief, each well was pulsed by addition of 10 μl WST-8 for Cell lines and cell culture. Human breast cancer cell lines BT549 and 2 h. Absorbance readings at a wavelength of 450nm were taken on Synergy H4 MCF7 were obtained from the cell bank of the Chinese Academy of Sciences Hybrid Microplate Reader. Breast cancer cells were plated on 6-cm dishes and were (Shanghai, China). MCF7 were cultured in Dulbecco’s modified Eagle’s medium counted every 2 days. Cells were stained with a 0.4% trypan blue solution and (Hyclone, Logan, UT, USA) with 10% FBS and 0.01 mg/ml human recombinant counted with a hemacytometer. For colony formation assay, 500 cells were placed in insulin. BT549 was maintained in RPMI 1640 (Hyclone) with 10% FBS. The cell line complete growth media and allowed to grow until visible colonies formed in a fresh MDA-231-D3H2LN (Xenogen, Alameda, CA, USA) was propagated in Minimum six-well plate (2 weeks). Cell colonies were fixed with cold methanol, stained with essential medium with Earl’s balanced salts solution (Hyclone) medium 0.1% crystal violet for 30 min, washed, air dried, photographed and counted. supplemented with 10% FBS, 1% non-essential amino acids (Hyclone) and 1% sodium pyruvate (Hyclone). Primary MEFs were prepared from littermate Anp32b+/+ Cell cycle analysis. For cell cycle analysis, all these BT549-transfected cells − − and Anp32b / E14.5 embroys. For transformed MEFs, primary MEFs were were pretreated with 2 mM thymidine twice to synchronize cells at G1/S border, and infected with a retrovirus generated from pLPC E1A/rasv12 using published then treated with 100 ng/ml nocodazole to block cells in G2/M for indicated times. To 29 6 techniques. All cells were fostered in a humidified atmosphere of 5% CO2. analyze cellular DNA content by flow cytometry, 10 cells were collected, rinsed and fixed overnight with 75% cold ethanol at − 20 °C. Cells were then treated with 100 μg/ml RNase A in Tris-HCl buffer (pH 7.4) and stained with 25 μg/ml propidium Patients. Fifty pairs of formalin-fixed and paraffin-embedded specimens of breast iodide. Samples were then subjected to the analysis by flow cytometry (FACSCalibur, cancer and adjacent normal tissues were purchased from Shanghai Outdo Biotech BD Biosciences, San Jose, CA, USA) using CellQuest Pro software (BD Biosciences). Co (Shanghai, China). Detailed information is described in the Supplementary Table Ten thousand cells were acquired and analyzed for the DNA content. S1. We obtained formalin-fixed and paraffin-embedded tumor specimens of breast cancer patients, which were histopathologically diagnosed during January 2003 and Animal experiments. Six-week-old female BALB/c nude mice were obtained June 2010 in the Department of Surgery, Shanghai First People’s Branch Hospital. from Shanghai SLAC Laboratory Animal Co., Shanghai, China. We subcutaneously All tumors were primary and were untreated before surgery. Detailed information is injected 2 × 106 tumor cell lines into left abdominal mammary fat pad. Starting described in the Supplementary Table S2. In addition, we also collected five pairs of 4 weeks post injection, the tumor size was monitored weekly by bioluminescence breast cancer and adjacent normal tissue specimens from Rui-Jin Hospital affiliated imaging. Mice were anesthetized each time and given intraperitoneal injection of to Shanghai Jiao-Tong University School of Medicine for analyzing ANP32B protein d-luciferin (150 μg/g body weight prepared in phosphate-buffered saline), and 10– expression. These studies were approved by the Medical Ethical Committee of the 15 min after the injection, bioluminescence images were captured with a charge- Affiliated Hospitals, Shanghai First People’s Branch Hospital and Rui-Jin Hospital, coupled device camera (IVIS; Xenogen). Mice were manipulated and housed respectively, and informed consent was obtained from all subjects or their relatives. according to protocols approved by Shanghai Medical Experimental Animal Care

Cell Death and Disease ANP32B deficiency suppresses proliferation and tumorigenesis S Yang et al 9

Commission. For the carcinogen DMBA treatment, virgin female Anp32b+/+ and 19. Kadkol SS, El Naga GA, Brody JR, Bai J, Gusev Y, Dooley WC et al. Expression of pp32 − − Anp32b / mice were treated with 1 mg doses of DMBA via oral gavage weekly for gene family members in breast cancer. Breast Cancer Res Treat 2001; 68:65–73. a total of 6 weeks. Mice were checked by palpation for tumor formation after 20. Tsukamoto Y, Uchida T, Karnan S, Noguchi T, Nguyen LT, Tanigawa M et al. Genome-wide completion of the DMBA treatments. analysis of DNA copy number alterations and gene expression in gastric cancer. J Pathol 2008; 216: 471–482. 21. Bjorck E, Ek S, Landgren O, Jerkeman M, Ehinger M, Bjorkholm M et al. High expression of Statistical analysis. All the statistical analyses were performed by the 105 – cyclin B1 predicts a favorable outcome in patients with follicular lymphoma. Blood 2005; : statistical package for social science (SPSS) (v.13) (SPSS Institute). The Mann 2908–2915. Whitney test was used to analyze the differences in different ANP32B expression 22. Reilly PT, Afzal S, Gorrini C, Lui K, Bukhman YV, Wakeham A et al. Acidic 2 groups. The χ -test was used to evaluate the correlation between ANP32B nuclear phosphoprotein 32kDa (ANP32)B-deficient mouse reveals a hierarchy expression and p-AKT expression. The P-values for comparison between line-linked of ANP32 importance in mammalian development. Proc Natl Acad Sci USA 2011; 108: groups were obtained by Student’s t-test. All statistical tests were two-sided, and 10243–10248. Po0.05 was considered to be statistically significant. 23. Opal P, Garcia JJ, McCall AE, Xu B, Weeber EJ, Sweatt JD et al. Generation and characterization of LANP/pp32 null mice. Mol Cell Biol 2004; 24: 3140–3149. 24. Reilly PT, Afzal S, Wakeham A, Haight J, You-Ten A, Zaugg K et al. Generation and Conflict of Interest characterization of the Anp32e-deficient mouse. PLoS One 2010; 5: e13597. The authors declare no conflict of interest. 25. Wong P, Leo VI, Low M, Mak TW, Zhang X, Reilly PT. Targeted ANP32E mutant mice do not demonstrate obvious movement defects. PLoS One 2013; 8: e63815. 26. Yu Y, Shen SM, Zhang FF, Wu ZX, Han B, Wang LS. Acidic leucine-rich nuclear Acknowledgements. This work was supported in part by grants from Ministry phosphoprotein 32 family member B (ANP32B) contributes to retinoic acid-induced differentiation of leukemic cells. Biochem Biophys Res Commun 2012; 423: 721–725. of Science and Technology (No. 2015CB910403, No. 2013CB910903), National 27. Xu Y, Liu J, Wu Y, Guo Q, Sun H, Chen G. Natural products against hemato- Natural Science Foundation (81230048; 81430061; 81502001, 91413115) and logical malignancies and identification of their targets. Sci China Life Sci 2015; 58: Natural Science Foundation from Science and Technology committee of Shanghai 1191–1201. (15ZR1426600). 28. Chakrabarti R, Hwang J, Andres Blanco M, Wei Y, Lukacisin M, Romano RA et al. Elf5 inhibits the epithelial-mesenchymal transition in mammary gland development and breast cancer metastasis by transcriptionally repressing Snail2. Nat Cell Biol 2012; 14: 1. Matilla A, Radrizzani M. The Anp32 family of proteins containing leucine-rich repeats. 1212–1222. Cerebellum 2005; 4:7–18. 29. Lin AW, Barradas M, Stone JC, van Aelst L, Serrano M, Lowe SW. Premature senescence 2. Seo SB, McNamara P, Heo S, Turner A, Lane WS, Chakravarti D. Regulation of histone involving p53 and p16 is activated in response to constitutive MEK/MAPK mitogenic acetylation and transcription by INHAT, a human cellular complex containing the set signaling. Genes Dev 1998; 12: 3008–3019. oncoprotein. Cell 2001; 104: 119–130. 30. Medina D, Lane HW, Shepherd F. Effect of dietary selenium levels on 7,12- 3. Seo SB, Macfarlan T, McNamara P, Hong R, Mukai Y, Heo S et al. Regulation of histone dimethylbenzanthracene-induced mouse mammary tumorigenesis. Carcinogenesis 1983; 4: acetylation and transcription by nuclear protein pp32, a subunit of the INHAT complex. J Biol 1159–1163. Chem 2002; 277: 14005–14010. 31. Eyster CA, Duggins QS, Olson AL. Expression of constitutively active Akt/protein kinase B 4. Obri A, Ouararhni K, Papin C, Diebold ML, Padmanabhan K, Marek M et al. ANP32E is a signals GLUT4 translocation in the absence of an intact actin cytoskeleton. J Biol Chem histone chaperone that removes H2A.Z from chromatin. Nature 2014; 505: 648–653. 2005; 280: 17978–17985. 5. Mao Z, Pan L, Wang W, Sun J, Shan S, Dong Q et al. Anp32e, a higher eukaryotic histone 32. Dillon RL, White DE, Muller WJ. The phosphatidyl inositol 3-kinase signaling network: chaperone directs preferential recognition for H2A.Z. Cell Res 2014; 24: 389–399. implications for human breast cancer. Oncogene 2007; 26: 1338–1345. 6. Tochio N, Umehara T, Munemasa Y, Suzuki T, Sato S, Tsuda K et al. Solution structure of 33. Skeen JE, Bhaskar PT, Chen CC, Chen WS, Peng XD, Nogueira V et al. Akt deficiency histone chaperone ANP32B: interaction with core histones H3-H4 through its acidic impairs normal cell proliferation and suppresses oncogenesis in a p53-independent and concave domain. J Mol Biol 2010; 401:97–114. mTORC1-dependent manner. Cancer Cell 2006; 10: 269–280. 7. Jiang X, Kim HE, Shu H, Zhao Y, Zhang H, Kofron J et al. Distinctive roles of PHAP proteins 34. Wood LD, Parsons DW, Jones S, Lin J, Sjoblom T, Leary RJ et al. The genomic landscapes and prothymosin-alpha in a death regulatory pathway. Science 2003; 299: 223–226. of human breast and colorectal cancers. Science 2007; 318: 1108–1113. 8. Shen SM, Yu Y, Wu YL, Cheng JK, Wang LS, Chen GQ. Downregulation of ANP32B, a novel 35. Dillon RL, Marcotte R, Hennessy BT, Woodgett JR, Mills GB, Muller WJ. Akt1 and akt2 play substrate of caspase-3, enhances caspase-3 activation and apoptosis induction in myeloid distinct roles in the initiation and metastatic phases of mammary tumor progression. Cancer leukemic cells. Carcinogenesis 2010; 31: 419–426. Res 2009; 69: 5057–5064. 9. Kim HE, Jiang X, Du F, Wang X. PHAPI, CAS, and Hsp70 promote apoptosome formation by 36. Zhou F, Li F, Xie F, Zhang Z, Huang H, Zhang L. TRAF4 mediates activation of TGF-beta preventing Apaf-1 aggregation and enhancing nucleotide exchange on Apaf-1. Mol Cell signaling and is a biomarker for oncogenesis in breast cancer. Sci China Life Sci 2014; 57: 2008; 30: 239–247. 1172–1176. 10. Radrizzani M, Vila-Ortiz G, Cafferata EG, Di Tella MC, Gonzalez-Guerrico A, Perandones C 37. Gao T, Furnari F, Newton AC. PHLPP: a phosphatase that directly dephosphorylates Akt, et al. Differential expression of CPD1 during postnatal development in the mouse cerebellum. promotes apoptosis, and suppresses tumor growth. Mol Cell 2005; 18:13–24. Brain Res 2001; 907: 162–174. 38. Liu CY, Hung MH, Wang DS, Chu PY, Su JC, Teng TH et al. Tamoxifen induces apoptosis 11. Li M, Makkinje A, Damuni Z. Molecular identification of I1PP2A, a novel potent heat-stable through cancerous inhibitor of protein phosphatase 2A-dependent phospho-Akt inacti- inhibitor protein of protein phosphatase 2A. Biochemistry 1996; 35: 6998–7002. vation in estrogen receptor-negative human breast cancer cells. Breast Cancer Res 2014; 12. Katayose Y, Li M, Al-Murrani SW, Shenolikar S, Damuni Z. Protein phosphatase 2A 16: 431. inhibitors, I(1)(PP2A) and I(2)(PP2A), associate with and modify the substrate specificity of 39. Toker A, Newton AC. Cellular signaling: pivoting around PDK-1. Cell 2000; 103: 185–188. protein phosphatase 1. J Biol Chem 2000; 275: 9209–9214. 40. Sansal I, Sellers WR. The biology and clinical relevance of the PTEN tumor suppressor 13. Brennan CM, Gallouzi IE, Steitz JA. Protein ligands to HuR modulate its interaction with pathway. J Clin Oncol 2004; 22: 2954–2963. target mRNAs in vivo. J Cell Biol 2000; 151:1–14. 14. Opal P, Garcia JJ, Propst F, Matilla A, Orr HT, Zoghbi HY. Mapmodulin/leucine-rich acidic nuclear protein binds the light chain of microtubule-associated protein 1B and modulates Cell Death and Disease is an open-access journal neuritogenesis. J Biol Chem 2003; 278: 34691–34699. published by Nature Publishing Group. This work is 15. Pan W, da Graca LS, Shao Y, Yin Q, Wu H, Jiang X. PHAPI/pp32 suppresses tumorigenesis licensed under a Creative Commons Attribution 4.0 International – by stimulating apoptosis. J Biol Chem 2009; 284: 6946 6954. License. The images or other third party material in this article are 16. Chen TH, Brody JR, Romantsev FE, Yu JG, Kayler AE, Voneiff E et al. Structure of pp32, an ’ acidic nuclear protein which inhibits oncogene-induced formation of transformed foci. Mol included in the article s Creative Commons license, unless indicated Biol Cell 1996; 7: 2045–2056. otherwise in the credit line; if the material is not included under the 17. Bai J, Brody JR, Kadkol SS, Pasternack GR. Tumor suppression and potentiation by Creative Commons license, users will need to obtain permission from manipulation of pp32 expression. Oncogene 2001; 20: 2153–2160. 18. Kadkol SS, Brody JR, Pevsner J, Bai J, Pasternack GR. Modulation of oncogenic potential by the license holder to reproduce the material. To view a copy of this alternative gene use in human prostate cancer. Nat Med 1999; 5: 275–279. license, visit http://creativecommons.org/licenses/by/4.0/

Supplementary Information accompanies this paper on Cell Death and Disease website (http://www.nature.com/cddis)

Cell Death and Disease